Cell, Vol. 67, 449-451,
November 1, 1991, Copyright 0 1991 by Cell Press
Minireview
Multiple Targets for Brefeldin A Hugh R. 6. Pelham Medical Research Council Laboratory of Molecular Biology Cambridge CB2 2QH England
Brefeldin A is a fungal metabolite that has profound and dramatic effects on the secretory pathway in mammalian cells (for review and references, see Wood et al., 1991; Lippincott-Schwartz et al., 1991; Hunziker et al., 1991). It not only inhibits secretion but also causes massive morphological changes: typically, the Golgi apparatus disintegrates and many Golgi enzymes are redistributed to the endoplasmic reticulum (ER). Remarkably, these changes are fully reversed when the drug is removed. These findings have generated excitement, because they suggest that an analysis of brefeldin action may help to reveal the general principles that govern the formation and maintenance of the organelles of the secretory pathway, and the rules that dictate the orderly traffic of transport vesicles between them. Recent results indicate that brefeldin has multiple, species-specific effects on vesicular transport that suggest several distinct sites of action in the endomembrane system. Principles of Vesicular Transport In recent years, many individual transport steps linking the ER, Golgi apparatus, endosomes, lysosomes, and plasma membrane have been documented or proposed (Figure 1). Movement of proteins between these organelles occurs by the budding and fusion of transport vesicles, although it is still unclear whether every compartment is a permanent structure, receiving and disgorging small vesicles, or whether the delivery and removal of selected components can convert, for example, early endosomes into late endo-
ER
Golgi
plasma I( membrane K \
lysosomes
4
t
TGN z
late endosomes
it f early endosomes
(transcytosis)
somes. In some cases, transport is clearly selective-a well-studied example being the preferential incorporation of cell surface receptors into endocytic vesicles (Pearse and Robinson, 1990). However, other steps seem less selective-newly synthesized membrane proteins that lack any obvious sorting signals are quite efficiently transported from the ER to the Golgi apparatus and from there to the cell surface. Transport vesicles must carry not only their cargo but also other molecules that specify their destination: in vitro experiments demonstrate that vesicles (and/or organelles) fuse only with appropriate target membranes (Goda and Pfeffer, 1969). Although it is not yet clear how this targeting is achieved, it seems likely that there are membrane proteins that travel with vesicles and mediate their interactions with acceptor membranes. The vesicular carriers are also likely to contain proteins that are attachment sites for microtubule motors such as kinesin and dynein (Vallee and Shpetner, 1990), since almost all transport steps involve, at least in some cell types, the microtubule-dependent movement of vesicles over relatively long intracellular distances. Microtubules are most obviously required for transport of vesicles along nerve axons, for transcytosis of endosome-derived vesicles across polarized ceils, and for the movement of endocytosed material to the Golgiproximal region. They may also be required for the movement of newly synthesized proteins from peripheral ER budding sites to the Golgi and from the Golgi to the plasma membrane, and for the recycling of endocytosed receptors to the leading edge of motile cells. In addition, microtubules mediate the spreading of ER tubules to the cell periphery and the accumulation of Golgi structures in the centrosomal region of the cell. There is no shortage of microtubule motors for these varied tasks-for example, Drosophila contains at least 11 and possibly as many as 35 different kinesin-related proteins (Goldstein, 1991)but to ensure appropriate vesicle movements and organelle locations, the attachment of these motors to membranes and/or their activity must be specifically and tightly regulated. Coat Proteins The molecules that are most likely to perform the crucial task of gathering together cargo proteins with the appropriate targeting molecules and motor attachment sites are the coat proteins, which cover the cytoplasmic face of most (if not all) budding vesicles. At the plasma membrane, endocytic receptors associate with proteins termed adaptins, which are themselves covered with a clathrin lattice. The adaptins are thought to provide the binding sites for the sorting signals that have been identified on the cytoplasmic tails of the receptors (Pearse and Robinson, 1990). A related but different adaptin complex is found on the surface of the trans-Golgi network, where it sorts mannose6-phosphate receptors into a different set of clathrin-coated vesicles destined for a pre-lysosomal compartment. Recently, a third coat complex has been identified on Golgi-derived vesicles. This coat does not include
Cell 450
NORMAL
BREFELDIN clathrin, but the one component whose sequence is known, termed p-COP, shows some homology to the adaptin family (Duden et al., i%i, Sttlaiull di ii., iL;Y i,. p-COP is found throughout the Golgl stack and may therefore be involved in several intercisternal transport steps. There are several transport steps for which no coat proteins have yet been identified (such as the recycling of material from endosomes to the plasma membrane, or its transcytosis to the apical membrane of epithelial cells), but there is no reason to believe that the list of coat proteins is complete. A simple view would be that each transport step (with the possible exception of the various intercisternal steps within the Golgi stack) uses a different coat protein complex. Membrane protems that are ~&&iir~~ transported would carry sorting signals recognized by the appropriate coats, and the different coat proteins would ultimately be responsible for the segregation of the intracellular membrane system into distinct organelles. Coat proteins may well have other important functions. By imposing curvature on a patch of membrane, they could
facilitate the formation and pinching off of vesicles (Lin et al., 1991). Moreover, by covering the (unidentified) membrane proteins that control the microtubule-dependent movement and fusion of vesicles, they could prevent the premature fusion of buds that have not yet been detached from the donor membrane (Figure 2). Premature fusion is an obvious potential problem in the Golgi stack, where donor and target membranes are in close proximity: buds presumably contain targeting molecules but must not undergo fusion, or adjacent cisternae would rapidly lose their individuality and coalesce into a single compartment. In vitro, the presence of coats does seem to prevent the fusion of Golgi-derived vesicles-a GTPyS-inhibitable uncoating step precedes fusion (Orci et al., 1989). For endocytic vesicles, at least, uncoating rapidly follows vesicle formation, suggesting that the task of the coat is complete once this has happened. Brefeldln Act/on on the Go/g/ Apparatus A key step in unraveling the complexities of brefeldin action came with the discovery that the drug causes a rapid loss of P-COP from Golgi membranes (Donaldson et al., 1990). In vitro, this dissociation is followed by the formation of continuous tubular connections between Golgi cisternae, as would be expected if one role of the coat is to prevent such events (Orci et al., 1991). In vivo, fusion of Golgi cisternae also seems to occur rapidly. In addition, long tubules emanate from the Golgi region, in a microtubule-dependent manner, and eventually fuse with the ER (Lippincott-Schwartz et al., 1990). Formation of these structures could be explained by the unmasking of microtubule motor attachment sites that are normally covered by coat proteins; motors would then pull out uncoated tubules, which would fail to pinch off into vesicles but would nevertheless be capable of fusion (Figure 2). A similar mechanism is presumably involved in the normal formation of tubules from the ER, Golgi apparatus, and endosomes. Usually, however, such tubules do not fuse with heterologous compartments. This view of brefeldin action differs somewhat from the original proposition that the tubular processes seen in the presence of brefeldin are the normal mediators of retrograde transport (Lippincott-Schwartz et al., 1990). It seems .., ’ ,~p\‘ b ; L;-.-,, i, -~pyh?nigme +hlt 379 qnqv+i_ >iiillt’ :I,\& y I::Ci: tally designed to prevent the formation of tubular connections between dissimilar organelles and the consequent intermingling of their contents, and that the structures and connections induced by brefeldin represent not a normal transport process, but aberrant events resulting from the breakdown of these mechanisms. Presumably, however, they involve the same targeting and fusion components that are utilized in vesicular transport. Brefeidin Action on Endosomes It has now been shown that brefeldin has effects not only svstem “I,. ,I.,,I” Cnlrri _Y.J’ ?nl..Icy __.._ r’_ , he,* c~rcnnn the enrlnsnmal and the transQolgi network (Wood et al., 1991; Lippincot&Schwartz et al., 1991; Hunziker et al., 1991). Since @COP is not found on endosomes, this suggests a second site of action for the drug, and this is borne out by the surprising observation that the Golgi and endosome effects show different species specificities: in rat and bovine
MInIreview 451
cells both organelles are affected, but in canine (MDCK) and kangaroo rat (PtK,) cells the Golgi complex is resistant while endosomes remain sensitive (Hunziker et al., 1991; Lippincott-Schwartz et al., 1991). The effect on MDCK cells is strikingly specific. Early endosomes form a tubular network in a microtubule-dependent fashion, but uptake of endocytic markers from the basolateral surface, recycling of receptors to the cell surface, and degradation of endocytosed material in (pre)lysosomes are essentially unaffected. Only transcytosis is inhibited, and Hunziker et al. (1991) show that it is the formation of transcytotic vesicles, rather than their subsequent transport and fusion, that is prevented. This clearly shows that the brefeldin effect is on a single transport step, not on a particular organelle. In the context of the model outlined above (and as suggested by Hunziker et al.), the results can easily be explained by the brefeldininduced dissociation of a transcytosis-specific coat protein from endosomes, which would prevent the formation of discrete transcytotic vesicles. The sorting and delivery of proteins to degradative compartments or to the basolateral surface would be unaffected, because these steps, although endosome-specific, would involve different coat proteins. In brefeldin-treated rat and bovine cells, endosomes have a slightly different fate: tubular networks extend from the transQolgi network (which is unaffected in MDCK cells) and fuse with early endosomes (Wood et al., 1991; Lippincott-Schwartz et al., 1991). This results in the formation of a mixed trans-Golgi network/endosomal compartment that is nevertheless still capable of receiving endocytosed material, recycling receptors to the cell surface, and delivering other molecules to a degradative compartment (this may not correspond to mature lysosomes, because transport to these does seem to be at least partially impaired). Interestingly, the late endosomelprelysosomal compartment remains distinct in at least some cell types (Wood et al., 1991) which suggests that late endosomes do not arise by maturation of early endosomes in these cells. Once again, brefeldin affects some transport steps but not others, which could be explained by the loss of some but not all coat proteins. The initial extension of trans-Golgi network tubules towards the periphery sugnatwnrk_tn_enrhcoma nath. oests that it ic 1 trans-Oolni way that IS prlrrlarlly allt?c;te~, II IdKll ~$j LI lt; LI dl I:, CUI~I I it;’ work-associated adaptinlclathrin coat complexes a likely site of action for the drug. Subsequently, the endosomel transQolgi network collapses into the centrosome region, perhaps driven by the motors that mediate the normal movement Of ep&~fyrn~l rn~t~rial to $0 virinitv of t;,c Golgi apparatus. Lippincott-Schwartz et al. (1991) show that even lysosomes are affected to some extent by brefeldin, forming a modest microtubule-dependent network. This suggests that some coat oroteins %sociatr? with them nnd raise@ the possibility that transport out of these organelles (which are usually considered a dead end) can occur. However, the lysosomes do not obviously fuse with any other organeke, so the destination of such traffic can only be guessed. The observation that brefeldin has multiple speciesspecific effects means that care must be taken in interpret-
ing its actions on poorly characterized systems. However, there does seem to be a common theme in all the results, which is consistent with the idea that the drug affects several different members, located in different organelles, of a family of related molecules- either coat proteins or some other components (perhaps GTP-binding proteins) involved in the assembly or disassembly of coats. It will be interesting to see whether other species reveal yet more patterns of brefeldin action, and whether such patterns will fit our current views of the mechanics and pathways of intracellular membrane traffic. References
Donaldson,J. G., Lippincott-Schwartz, J., and Klausner,
Bloom, G. Et., Kreis, T. E., Ft. D. (1990). J. Cell Biol. 777, 2295-2306.
Duden, Ft., Griffiths. G., Frank, Ft., Argo% P., and Kreis, T. E. (1991). Cell 64, 649-665. Goda, Y., and Pfeffer, S. R. (1969). FASEB J. 3, 2466-2495. Goldstein,
L. S. B. (1991). Trends Cell Biol. 1, 93-96
Hunziker, W., Whitney, J. A., and Mellman, I. (1991). Cell, this issue. Lin, H. C., Moore, M. S., Sanan. D. A., and Anderson, J. Cell Biol. 774, 861-691.
R. G. W. (1991).
Lippincott-Schwartz. J., Donaldson, J. G., Schweizer. A.. Berger, E.G., Hauri. H. P., Yuan, L. C., and Klausner, R. D. (1990). Cell 60, 621-636. Lippincott-Schwartz. J., Yuan, L., Tipper, C.. Amherdt, and Klausner, R. D. (1991). Cell, this issue. Orci, L., Malhotra, V., Amherdt, (1969). Cell 56. 357-366.
M., Orci, L.,
hf., Serafini, T., and Rothman, J. E.
Orci, L., Tagaya, M., Amherdt, Lippincott-Schwartz, J., Klausner, Cell 64, 1163-1195.
M., Perrelet, A., Donaldson, J., R. D., and Rothman, J. E. (1991).
Pearse, B. M. F., and Robinson, M. S. (1990). Annu. Rev. Cell Biol. 6, 151-171. Serafini, T., Stenbeck, G., Brecht, A.. Lottspeich, F.. Orci, L., Rothman, J. E., and Wieland, F. T. (1991). Nature 349, 215-220. Vallee, R. B., and Shpetner. 909-932.
H. S. (1990). Annu. Rev. B&hem.
59,
Wood, S. A., Park, J. E., and Brown, W. J. (1991). Cell, this issue.
November 1, 1991, Copyright 0 1991 by Cell Press
Minireview
Multiple Targets for Brefeldin A Hugh R. 6. Pelham Medical Research Council Laboratory of Molecular Biology Cambridge CB2 2QH England
Brefeldin A is a fungal metabolite that has profound and dramatic effects on the secretory pathway in mammalian cells (for review and references, see Wood et al., 1991; Lippincott-Schwartz et al., 1991; Hunziker et al., 1991). It not only inhibits secretion but also causes massive morphological changes: typically, the Golgi apparatus disintegrates and many Golgi enzymes are redistributed to the endoplasmic reticulum (ER). Remarkably, these changes are fully reversed when the drug is removed. These findings have generated excitement, because they suggest that an analysis of brefeldin action may help to reveal the general principles that govern the formation and maintenance of the organelles of the secretory pathway, and the rules that dictate the orderly traffic of transport vesicles between them. Recent results indicate that brefeldin has multiple, species-specific effects on vesicular transport that suggest several distinct sites of action in the endomembrane system. Principles of Vesicular Transport In recent years, many individual transport steps linking the ER, Golgi apparatus, endosomes, lysosomes, and plasma membrane have been documented or proposed (Figure 1). Movement of proteins between these organelles occurs by the budding and fusion of transport vesicles, although it is still unclear whether every compartment is a permanent structure, receiving and disgorging small vesicles, or whether the delivery and removal of selected components can convert, for example, early endosomes into late endo-
ER
Golgi
plasma I( membrane K \
lysosomes
4
t
TGN z
late endosomes
it f early endosomes
(transcytosis)
somes. In some cases, transport is clearly selective-a well-studied example being the preferential incorporation of cell surface receptors into endocytic vesicles (Pearse and Robinson, 1990). However, other steps seem less selective-newly synthesized membrane proteins that lack any obvious sorting signals are quite efficiently transported from the ER to the Golgi apparatus and from there to the cell surface. Transport vesicles must carry not only their cargo but also other molecules that specify their destination: in vitro experiments demonstrate that vesicles (and/or organelles) fuse only with appropriate target membranes (Goda and Pfeffer, 1969). Although it is not yet clear how this targeting is achieved, it seems likely that there are membrane proteins that travel with vesicles and mediate their interactions with acceptor membranes. The vesicular carriers are also likely to contain proteins that are attachment sites for microtubule motors such as kinesin and dynein (Vallee and Shpetner, 1990), since almost all transport steps involve, at least in some cell types, the microtubule-dependent movement of vesicles over relatively long intracellular distances. Microtubules are most obviously required for transport of vesicles along nerve axons, for transcytosis of endosome-derived vesicles across polarized ceils, and for the movement of endocytosed material to the Golgiproximal region. They may also be required for the movement of newly synthesized proteins from peripheral ER budding sites to the Golgi and from the Golgi to the plasma membrane, and for the recycling of endocytosed receptors to the leading edge of motile cells. In addition, microtubules mediate the spreading of ER tubules to the cell periphery and the accumulation of Golgi structures in the centrosomal region of the cell. There is no shortage of microtubule motors for these varied tasks-for example, Drosophila contains at least 11 and possibly as many as 35 different kinesin-related proteins (Goldstein, 1991)but to ensure appropriate vesicle movements and organelle locations, the attachment of these motors to membranes and/or their activity must be specifically and tightly regulated. Coat Proteins The molecules that are most likely to perform the crucial task of gathering together cargo proteins with the appropriate targeting molecules and motor attachment sites are the coat proteins, which cover the cytoplasmic face of most (if not all) budding vesicles. At the plasma membrane, endocytic receptors associate with proteins termed adaptins, which are themselves covered with a clathrin lattice. The adaptins are thought to provide the binding sites for the sorting signals that have been identified on the cytoplasmic tails of the receptors (Pearse and Robinson, 1990). A related but different adaptin complex is found on the surface of the trans-Golgi network, where it sorts mannose6-phosphate receptors into a different set of clathrin-coated vesicles destined for a pre-lysosomal compartment. Recently, a third coat complex has been identified on Golgi-derived vesicles. This coat does not include
Cell 450
NORMAL
BREFELDIN clathrin, but the one component whose sequence is known, termed p-COP, shows some homology to the adaptin family (Duden et al., i%i, Sttlaiull di ii., iL;Y i,. p-COP is found throughout the Golgl stack and may therefore be involved in several intercisternal transport steps. There are several transport steps for which no coat proteins have yet been identified (such as the recycling of material from endosomes to the plasma membrane, or its transcytosis to the apical membrane of epithelial cells), but there is no reason to believe that the list of coat proteins is complete. A simple view would be that each transport step (with the possible exception of the various intercisternal steps within the Golgi stack) uses a different coat protein complex. Membrane protems that are ~&&iir~~ transported would carry sorting signals recognized by the appropriate coats, and the different coat proteins would ultimately be responsible for the segregation of the intracellular membrane system into distinct organelles. Coat proteins may well have other important functions. By imposing curvature on a patch of membrane, they could
facilitate the formation and pinching off of vesicles (Lin et al., 1991). Moreover, by covering the (unidentified) membrane proteins that control the microtubule-dependent movement and fusion of vesicles, they could prevent the premature fusion of buds that have not yet been detached from the donor membrane (Figure 2). Premature fusion is an obvious potential problem in the Golgi stack, where donor and target membranes are in close proximity: buds presumably contain targeting molecules but must not undergo fusion, or adjacent cisternae would rapidly lose their individuality and coalesce into a single compartment. In vitro, the presence of coats does seem to prevent the fusion of Golgi-derived vesicles-a GTPyS-inhibitable uncoating step precedes fusion (Orci et al., 1989). For endocytic vesicles, at least, uncoating rapidly follows vesicle formation, suggesting that the task of the coat is complete once this has happened. Brefeldln Act/on on the Go/g/ Apparatus A key step in unraveling the complexities of brefeldin action came with the discovery that the drug causes a rapid loss of P-COP from Golgi membranes (Donaldson et al., 1990). In vitro, this dissociation is followed by the formation of continuous tubular connections between Golgi cisternae, as would be expected if one role of the coat is to prevent such events (Orci et al., 1991). In vivo, fusion of Golgi cisternae also seems to occur rapidly. In addition, long tubules emanate from the Golgi region, in a microtubule-dependent manner, and eventually fuse with the ER (Lippincott-Schwartz et al., 1990). Formation of these structures could be explained by the unmasking of microtubule motor attachment sites that are normally covered by coat proteins; motors would then pull out uncoated tubules, which would fail to pinch off into vesicles but would nevertheless be capable of fusion (Figure 2). A similar mechanism is presumably involved in the normal formation of tubules from the ER, Golgi apparatus, and endosomes. Usually, however, such tubules do not fuse with heterologous compartments. This view of brefeldin action differs somewhat from the original proposition that the tubular processes seen in the presence of brefeldin are the normal mediators of retrograde transport (Lippincott-Schwartz et al., 1990). It seems .., ’ ,~p\‘ b ; L;-.-,, i, -~pyh?nigme +hlt 379 qnqv+i_ >iiillt’ :I,\& y I::Ci: tally designed to prevent the formation of tubular connections between dissimilar organelles and the consequent intermingling of their contents, and that the structures and connections induced by brefeldin represent not a normal transport process, but aberrant events resulting from the breakdown of these mechanisms. Presumably, however, they involve the same targeting and fusion components that are utilized in vesicular transport. Brefeidin Action on Endosomes It has now been shown that brefeldin has effects not only svstem “I,. ,I.,,I” Cnlrri _Y.J’ ?nl..Icy __.._ r’_ , he,* c~rcnnn the enrlnsnmal and the transQolgi network (Wood et al., 1991; Lippincot&Schwartz et al., 1991; Hunziker et al., 1991). Since @COP is not found on endosomes, this suggests a second site of action for the drug, and this is borne out by the surprising observation that the Golgi and endosome effects show different species specificities: in rat and bovine
MInIreview 451
cells both organelles are affected, but in canine (MDCK) and kangaroo rat (PtK,) cells the Golgi complex is resistant while endosomes remain sensitive (Hunziker et al., 1991; Lippincott-Schwartz et al., 1991). The effect on MDCK cells is strikingly specific. Early endosomes form a tubular network in a microtubule-dependent fashion, but uptake of endocytic markers from the basolateral surface, recycling of receptors to the cell surface, and degradation of endocytosed material in (pre)lysosomes are essentially unaffected. Only transcytosis is inhibited, and Hunziker et al. (1991) show that it is the formation of transcytotic vesicles, rather than their subsequent transport and fusion, that is prevented. This clearly shows that the brefeldin effect is on a single transport step, not on a particular organelle. In the context of the model outlined above (and as suggested by Hunziker et al.), the results can easily be explained by the brefeldininduced dissociation of a transcytosis-specific coat protein from endosomes, which would prevent the formation of discrete transcytotic vesicles. The sorting and delivery of proteins to degradative compartments or to the basolateral surface would be unaffected, because these steps, although endosome-specific, would involve different coat proteins. In brefeldin-treated rat and bovine cells, endosomes have a slightly different fate: tubular networks extend from the transQolgi network (which is unaffected in MDCK cells) and fuse with early endosomes (Wood et al., 1991; Lippincott-Schwartz et al., 1991). This results in the formation of a mixed trans-Golgi network/endosomal compartment that is nevertheless still capable of receiving endocytosed material, recycling receptors to the cell surface, and delivering other molecules to a degradative compartment (this may not correspond to mature lysosomes, because transport to these does seem to be at least partially impaired). Interestingly, the late endosomelprelysosomal compartment remains distinct in at least some cell types (Wood et al., 1991) which suggests that late endosomes do not arise by maturation of early endosomes in these cells. Once again, brefeldin affects some transport steps but not others, which could be explained by the loss of some but not all coat proteins. The initial extension of trans-Golgi network tubules towards the periphery sugnatwnrk_tn_enrhcoma nath. oests that it ic 1 trans-Oolni way that IS prlrrlarlly allt?c;te~, II IdKll ~$j LI lt; LI dl I:, CUI~I I it;’ work-associated adaptinlclathrin coat complexes a likely site of action for the drug. Subsequently, the endosomel transQolgi network collapses into the centrosome region, perhaps driven by the motors that mediate the normal movement Of ep&~fyrn~l rn~t~rial to $0 virinitv of t;,c Golgi apparatus. Lippincott-Schwartz et al. (1991) show that even lysosomes are affected to some extent by brefeldin, forming a modest microtubule-dependent network. This suggests that some coat oroteins %sociatr? with them nnd raise@ the possibility that transport out of these organelles (which are usually considered a dead end) can occur. However, the lysosomes do not obviously fuse with any other organeke, so the destination of such traffic can only be guessed. The observation that brefeldin has multiple speciesspecific effects means that care must be taken in interpret-
ing its actions on poorly characterized systems. However, there does seem to be a common theme in all the results, which is consistent with the idea that the drug affects several different members, located in different organelles, of a family of related molecules- either coat proteins or some other components (perhaps GTP-binding proteins) involved in the assembly or disassembly of coats. It will be interesting to see whether other species reveal yet more patterns of brefeldin action, and whether such patterns will fit our current views of the mechanics and pathways of intracellular membrane traffic. References
Donaldson,J. G., Lippincott-Schwartz, J., and Klausner,
Bloom, G. Et., Kreis, T. E., Ft. D. (1990). J. Cell Biol. 777, 2295-2306.
Duden, Ft., Griffiths. G., Frank, Ft., Argo% P., and Kreis, T. E. (1991). Cell 64, 649-665. Goda, Y., and Pfeffer, S. R. (1969). FASEB J. 3, 2466-2495. Goldstein,
L. S. B. (1991). Trends Cell Biol. 1, 93-96
Hunziker, W., Whitney, J. A., and Mellman, I. (1991). Cell, this issue. Lin, H. C., Moore, M. S., Sanan. D. A., and Anderson, J. Cell Biol. 774, 861-691.
R. G. W. (1991).
Lippincott-Schwartz. J., Donaldson, J. G., Schweizer. A.. Berger, E.G., Hauri. H. P., Yuan, L. C., and Klausner, R. D. (1990). Cell 60, 621-636. Lippincott-Schwartz. J., Yuan, L., Tipper, C.. Amherdt, and Klausner, R. D. (1991). Cell, this issue. Orci, L., Malhotra, V., Amherdt, (1969). Cell 56. 357-366.
M., Orci, L.,
hf., Serafini, T., and Rothman, J. E.
Orci, L., Tagaya, M., Amherdt, Lippincott-Schwartz, J., Klausner, Cell 64, 1163-1195.
M., Perrelet, A., Donaldson, J., R. D., and Rothman, J. E. (1991).
Pearse, B. M. F., and Robinson, M. S. (1990). Annu. Rev. Cell Biol. 6, 151-171. Serafini, T., Stenbeck, G., Brecht, A.. Lottspeich, F.. Orci, L., Rothman, J. E., and Wieland, F. T. (1991). Nature 349, 215-220. Vallee, R. B., and Shpetner. 909-932.
H. S. (1990). Annu. Rev. B&hem.
59,
Wood, S. A., Park, J. E., and Brown, W. J. (1991). Cell, this issue.
